In the context of our J-Cell project, we propose a novel approach for locating objects in a fully decentralized system. In our system migrations are caused by processor nodes modifying shared objects. Hence our system...
详细信息
Load balancing increases the efficient use of existing resources for parallel and distributed applications, At a coarse level of granularity, advances in runtime systems for parallel programs have been proposed in ord...
详细信息
Load balancing increases the efficient use of existing resources for parallel and distributed applications, At a coarse level of granularity, advances in runtime systems for parallel programs have been proposed in order to control available resources as efficiently as possible by utilizing idle resources and using task migration. Simultaneously, at a finer granularity level, advances in algorithmic strategies for dynamically balancing computational loads by data redistribution have been proposed in order to respond to variations in processor performance during the execution of a given parallel application Combining strategies from each level of granularity can result in a system which delivers advantages of both. The resulting integration is systemic in nature and transfers the responsibility of efficient resource utilization from the application programmer to the runtime system. This paper presents the design and implementation of a system that combines an algorithmic fine-grained data parallel load balancing strategy with a systemic coarse-grained task-parallel load balancing strategy, and reports on recent experimental results of running a computationally intensive scientific application under this integrated system. The experimental results indicate that a distributed runtime environment which combines both task and data migration can provide performance advantages with little overhead. It also presents proposals for performance enhancements of the implementation, as well as future explorations for effective resource management. Copyright (C) 2001 John Wiley & Sons, Ltd.
Todays situation is characterized by an increasing pervasiveness of a plethora of mobile devices featuring different capabilities and exhibiting different system interfaces making the handling of these devices and esp...
详细信息
ISBN:
(纸本)9780769551593
Todays situation is characterized by an increasing pervasiveness of a plethora of mobile devices featuring different capabilities and exhibiting different system interfaces making the handling of these devices and especially the cooperation between different devices a complex task. In this paper, we consider the sum of all these devices as one emerging system (the swarm) and present an approach of a swarm operating system that on a systemic level manages these devices (local devices give up their autonomy) while providing a common interface to user applications. We provide a programming model for distributed mobile applications that abstracts from error-prone aspects such as distribution and concurrency by giving the programmer a systemic view to system resources. The model allows the programmer to define actions that can be restricted in space and time. Together with a high level goal, an entire application emerges implicitly based on those defined actions. In order to execute such applications, we present an architecture for a runtime system that uses virtualization techniques in order to execute multiple independently developed applications in parallel. The system follows a service-oriented architecture: one of the core services is the space-time scheduler that plans applications (a set of actions) in time and space.
暂无评论